C3-Positive Reactive Astrocyte Induction Increases Primary Cilium Length

While it has been reported that brain astrocytes can assemble the primary cilium, its physiological functions remain largely unexplored. Additionally, although there have been numerous suggestions regarding the role of the primary cilium in immune response regulation, detailed analyses are lacking. To investigate the physiological functions of the primary cilium in neurotoxic C3-positive reactive astrocytes, we isolated glial cells from the brains of postnatal day 7 (P7) mice and induced C3-positive reactive astrocyte differentiation by stimulating them with LPS (Fig. 1A) (Liddelow et al. 2017). Subsequently, we observed the expression of the C3-positive reactive astrocyte marker C3 and the primary cilium in cultured astrocytes. Stimulation with LPS in mixed glial cell cultures upregulated C3 expression in astrocytes (Fig. S1A, B). Under these conditions, more than 90% of the astrocytes formed primary cilia, and the primary cilium formation rate remained unchanged following LPS stimulation (Fig. 1B, C). However, LPS stimulation in mixed glial cell cultures significantly increased the length of the primary cilium in astrocytes (Fig. 1B, D).

Fig. 1

C3-positive reactive astrocyte induction increases primary cilium length. A Schematic illustration of C3-positive reactive astrocyte induction. Microglial activation by LPS stimulation upregulates the expression of IL-1α, C1q, and TNF-α, increasing the expression of C3 in reactive astrocytes (Liddelow et al. 2017). Figures were created with BioRender.com. B Representative immunostaining image of Arl13B (primary cilium, green) in GFAP-positive cells (astrocytes) (red). Mixed cortical glial cells were treated with 100 ng/ml LPS for 24, 48, and 72 h. The primary cilium indicated by arrow was magnified and shown on the left bottom. Nuclei were stained with Hoechst 33342 (blue). Scale bar; 20 μm. Four independent experiments were performed. C Percentage of astrocytes (GFAP+ARL13B+ cells/GFAP+ cells) with a primary cilium shown Fig. 1B shows in the graph. Glial cells were stimulated with PBS or 100 ng/ml LPS for 24, 48, 72 h. Four independent experiments were performed (PBS 24 h; 107 ciliated cells/259 GFAP+ cells, LPS 24 h; 102 ciliated cells/253 GFAP+ cells, PBS 48 h; 136 ciliated cells/290 GFAP+ cells, LPS 48 h; 121 ciliated cells/270 GFAP+ cells, PBS 72 h; 145 ciliated cells/347 GFAP+ cells, LPS 72 h; 96 ciliated cells/236 GFAP+ cells). The Mann‒Whitney U test was performed. ns, nonsignificant. Error bar indicates SD. D The astrocytic primary cilium length shown in Fig. 1B was measured (PBS 24 h; 108 cilia, LPS 24 h; 92 cilia, PBS 48 h; 107 cilia, LPS 48 h; 97 cilia, PBS 72 h; 93 cilia, LPS 72 h; 84 cilia). Four independent experiments were performed. Error bar indicates SD. ***P < 0.001, ****P < 0.0001. (Mann‒Whitney U test). E Representative image of Arl13B (red) and C3 (green) and nuclei (Hoechst 33342, blue) in enriched astrocytes. Astrocytes were enriched from glial mixture culture and stimulated with PBS or a cytokine mixture (3 ng/mL IL-1α, 30 ng/ml TNFα, 400 ng/ml C1q) for 24 h. The arrow indicates the primary cilium. Scale bar; 5 μm. Four independent experiments were performed. F (left) Percentage of C3-positive cells (C3+ cells/total cells) in enriched astrocytes culture shown in Fig. 1E. Four independent experiments were performed. (PBS; 46 C3+ cells /779 total cells, Cyt; 293 C3+ cells /783 total cells) Error bar indicates SD. *P < 0.05, (Mann‒Whitney U test). (Right) Astrocytic cilia length is shown in the graph (PBS; 177 cilia, Cyt; 175 cilia). Four independent experiments were performed. Error bar indicates SD. ****P < 0.0001. (Mann‒Whitney U test). G The correlation coefficient between the expression level of C3 and the length of cilium. Primary cilium length and C3 expression intensity was measured by Image J. Images of astrocytes stimulated with PBS or cytokines were randomly selected, and five images were chosen for each condition. Quantification was performed for all cells within the selected images

It has been reported that C3-positive reactive astrocytes are induced in response to three cytokines secreted by activated microglia, IL-1⍺, TNF-⍺, and C1q (Liddelow et al. 2017) (Fig. 1A). Therefore, we enriched astrocytes from glial cultures, and stimulated them with these cytokines and observed the primary cilium after stimulation (Fig. 1A). Please note that 95% or more of the enriched cells were GFAP-positive. As expected from the results obtained using mixed glial cell cultures, the cytokine stimulation of enriched astrocytes increased the proportion of C3-positive cells, accompanied by increased primary cilium length (Fig. 1E, F). Moreover, when we calculated the correlation coefficient between the length of primary cilium and the expression level of C3, the R2-value was 0.2471 (Fig. 1G). To further investigate the correlation between C3-positive reactive astrocyte differentiation and primary cilium length, we examined the time it took for the expression level of C3 and cilium length to increase following cytokine stimulation. It was suggested that both C3-positive cell numbers and primary cilium length gradually increased after cytokine stimulation, starting from 1 h poststimulation (Fig. S1C, D). These results suggest that LPS activates microglia, promoting cytokine production and C3 expression in reactive astrocyte, accompanied by primary cilium elongation.

TRPV4 Activation Induces C3-Positive Reactive Astrocyte

Transient receptor potential vanilloid 4 (TRPV4) is an ion channel localized in the primary cilium, and intraperitoneal (i.p.) administration of the TRPV4 activator GSK1016790A enhanced the activation of microglia and astrocytes in the hippocampal region in mice (Wang et al. 2019). Since a specific population of astrocytes functionally express TRPV4 (Shibasaki et al. 2014), we analysed whether the activation of TRPV4 in astrocytes affected the expression of C3 in reactive astrocytes. GSK1016790A, a TRPV4 agonist, promoted the expression of C3 in reactive astrocytes and elongated the primary cilium, which was similar to the effects observed in response to cytokine stimulation. (Fig. 2A–E). In contrast, treatment with the TRPV4 antagonist GSK2193874 attenuated the effects of GSK1016790A although statistical analysis did not show the significant difference. Interestingly, cytokine-induced expression of C3 in reactive astrocytes was also attenuated by GSK2193874 (Fig. 2A, C).

Fig. 2

TRPV4 activation induces C3-positive reactive astrocyte. A Representative image of C3 (green), GFAP (red), and Hoechst 33342 (blue) in enriched astrocytes. Cells were stimulated with PBS (negative control), a cytokine mixture (Cyt) (3 ng/mL IL-1α, 30 ng/ml TNFα, 400 ng/ml C1q), 100 nM GSK1016790A (GSK101), 100 nM GSK2193874 (GSK219) for 24 h. Scale bar; 20 µm. n = 5 independent experiments B Representative image of Arl13B (green) and Hoechst 33342 (blue) in enriched astrocytes. Cells were stimulated with a cytokine mixture (Cyt) (3 ng/mL IL-1α, 30 ng/ml TNFα, 400 ng/ml C1q), 100 nM GSK1016790A (GSK101), 100 nM GSK2193874 (GSK219) for 24 h. The primary cilium indicated by arrow was magnified and shown on the right bottom. n = 4 independent experiments. Scale bar; 20 µm. C The percentage of C3-positive astrocytes (C3+GFAP+ cells/GFAP+ cells) shown Fig. 2A shows in the graph (PBS; 76 C3+ GFAP+ cells/484 GFAP+ cells, Cyt; 244 C3+ GFAP+ cells/378 GFAP+ cells, GSK101; 247 C3+ GFAP+ cells/408 GFAP+ cells, GSK219; 55 C3+ GFAP+ cells/373 GFAP+ cells, Cyt + GSK101; 268 C3+ cells/439 total cells, Cyt + GSK219; 99 C3+ cells/321 total cells, GSK101 + GSK219; 94 C3+ GFAP+ cells/ 409 GFAP+ cells, Cyt + GSK101 + GSK219; 164 C3+ GFAP+ cells/399 GFAP+ cells). *P < 0.05, **P < 0.01, ns, nonsignificant. (Kruskal–Wallis test, Dunn’s multiple comparison). n = 5 independent experiments. Error bar indicates SD. D Cilium length shown in Fig. 2B shows in graph (ctrl; 176 cilia, Cyt; 141 cilia, GSK101; 151 cilia, GSK219; 210 cilia). **P < 0.01, ****P < 0.0001, ns, nonsignificant (Kruskal–Wallis test, Dunn’s multiple comparison). Error bar indicates SD. n = 4 independent experiments. E The percentage of ciliated astrocytes (Arl13B+ GFAP+ cells/GFAP+ cells) shown Fig. 2B shows in the graph (ctrl; 190 ciliated cells/308 GFAP+ cells, Cyt; 159 ciliated cells/200 GFAP+ cells, GSK101; 169 ciliated cells/198 GFAP+ cells, GSK219; 228 ciliated cells/278 GFAP+ cells). ns, nonsignificant (Kruskal–Wallis test, Dunn’s multiple comparison). n = 4 independent experiments. Error bar indicates SD. F Representative image of C3 (green), GFAP (red), and Hoechst 33342 (blue) in purified primary astrocytes. Cells were stimulated with PBS (negative control), a cytokine mixture (Cyt) (3 ng/mL IL-1α, 30 ng/ml TNFα, 400 ng/ml C1q), 100 nM GSK1016790A (GSK101), 100 nM UTP, 100 nM YM-254890 (YM-254) for 24 h. n = 5 independent experiments. Scale bar; 20 µm. G The percentage of C3-positive astrocytes (C3+ GFAP+ cells/GFAP+ cells) shown Fig. 2F shows in the graph (PBS; 23 C3+ GFAP+ cells/278 GFAP+ cells, Cyt; 138 C3+ GFAP+ cells/294 GFAP+ cells, GSK101; 117 C3+ GFAP+ cells/ 312 GFAP+ cells, UTP; 152 C3+ GFAP+ cells/305 GFAP+ cells, YM-254; 23 C3+ GFAP+ cells/278 GFAP+ cells, Cyt + YM-254; 54 C3+ GFAP+ cells/321 GFAP+ cells, GSK101 + YM-254; 41 C3+ GFAP+ cells/329 GFAP+ cells, UTP + YM-254; 32 C3+ GFAP+ cells/270 GFAP+ cells). **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, nonsignificant. (Kruskal–Wallis test, Dunn’s multiple comparison). Error bar indicates SD. n = 5 independent experiments. H Representative image of C3 (green), GFAP (red), and Hoechst 33342 (blue) in enriched astrocytes. Cells were stimulated with a cytokine mixture (Cyt) (3 ng/mL IL-1α, 30 ng/ml TNFα, 400 ng/ml C1q), 100 nM GSK1016790A (GSK101) and 1 µM BAPTA-AM for 24 h. n = 5 independent experiments. Scale bar; 20 µm. I The percentage of C3-positive astrocytes (C3+ GFAP+ cells/GFAP+ cells) shown Fig. 2H shows in the graph (PBS, BAPTA-; 17 C3+ GFAP+ cells/354 GFAP+ cells, PBS, BAPTA + ; 18 C3+ GFAP+ cells/318 GFAP+ cells, Cyt, BAPTA-; 128 C3+ GFAP+ cells/ 275 GFAP+ cells, Cyt, BAPTA + ; 27 C3+ GFAP+ cells/192 GFAP+ cells, GSK101, BAPTA-; 127 C3+ cells/270 total cells, GSK101, BAPTA + ; 49 C3+ cells/313 total cells). *P < 0.05, **P < 0.01, ns, nonsignificant (The Mann‒Whitney U test). n = 5 independent experiments

Activation of TRPV4 in astrocytes has been reported to promote ATP and glutamate release, which induces reactive astrocyte (Shibasaki et al. 2014). Since both ATP and glutamate are ligands for G protein-coupled receptors (GPCRs), we hypothesized that the activation of TRPV4 may lead to secondary activation of GPCRs, including P2Y receptors, leading to C3 expression in reactive astrocyte. To investigate whether the pharmacological modulation of GPCR-mediated signalling affects C3 expression in reactive astrocyte differentiation, we stimulated astrocytes with 100 nM UTP, a P2Y receptor ligand. We found that UTP increased the number of C3-positive reactive astrocytes (Fig. 2F, G). Furthermore, 100 nM YM-254890, an inhibitor of the trimeric G protein Gq, did not increase the number of C3-positive reactive astrocytes even when the astrocytes were stimulated by cytokines, GSK1016790A and UTP (Fig. 2F, G). These results suggest the activation of secondary GPCRs following TRPV4 activation, which promotes C3 expression in reactive astrocyte.

Since the activation of TRPV4 promotes C3 expression in reactive astrocyte, we performed an analysis using the intracellular calcium chelator BAPTA-AM to investigate whether the influx of Ca2+ through TRPV4 regulated C3 expression in reactive astrocyte. We observed that the increase in C3-positive reactive astrocytes induced by GSK1016790A was decreased by BAPTA-AM treatment (Fig. 2H, I). Additionally, the number of C3-positive reactive astrocytes increased by cytokine stimulation was decreased by BAPTA-AM treatment (Fig. 2H, I). These results suggest that the promotion of C3 expression in reactive astrocyte induced by TRPV4 activation is mediated by the increase in intracellular calcium ions.

Inflammation Model Mice Exhibit an Increase in C3-Positive Reactive Astrocyte Numbers and Elongation of the Primary Cilium in the Brain

To observe the changes in C3-positive reactive astrocytes numbers and primary cilium length in the brains of adult mice, we established an inflammation model by intraperitoneally (i.p.) administering LPS. In the inflammation model mice, the number of C3-positive astrocytes were increased compared that in the control group treated with PBS, which was in accordance with previous findings (Liddelow et al. 2017) (Fig. S2A, B). When observing the primary cilium, we found that the length of primary cilia in both astrocytes and C3-positive cells was significantly increased compared to that in the control group (Figs. 3A, B, S2C, D). The primary cilium formation rate remained unchanged in both astrocytes and non-astrocytic cells (Fig. 3B, S2D). These results suggest that inflammation induced by i.p. LPS administration elongates the primary cilium in astrocytes in the brain.

Fig. 3

LPS injection elongates the astrocytic primary cilium in mice brain. A Representative image of GFAP (astrocytes, red), Arl13B (primary cilium, green), and nuclei (blue) in the mouse hippocampal area. Three-month-old mice were i.p. injected with 1 mg/kg LPS twice a week for 6 weeks. Scale bar; 10 μm. PBS-treated mice; n = 5 (biological replicates). LPS-treated mice; n = 5 (biological replicates). B The percentage of ciliated astrocytes (Arl13B+ GFAP+ cells/GFAP+ cells) (left, PBS; 75 ciliated cells /521 GFAP+ cells, LPS; 36 ciliated cells /382 GFAP+ cells), percentage of ciliated GFAP-negative cells (Arl13B+ GFAP− cells/GFAP− cells) (middle, PBS; 31 ciliated cells /341 GFAP− cells, LPS; 20 ciliated cells /209 GFAP− cells), and astrocytic primary cilium length (right, PBS; 103 primary cilia, LPS; 36 primary cilia) shown in Fig. 3A are shown in the graph. ****P < 0.0001, ns, nonsignificant. (Mann‒Whitney U test). Five biological replicates (5 mice) were examined respectively. Error bar indicates SD

The Specific Inhibition of Primary Cilium Formation in Astrocytes Decreases C3 Expression in Reactive Astrocyte

IFT88 is essential for primary cilium formation and maintenance, and its downregulation has been known to cause primary cilium loss or shortening (Fliegauf et al. 2007; Valente et al. 2014). To investigate the physiological function of the primary cilium in C3 expression in reactive astrocyte, we transfected siRNA targeting IFT88 into cultured astrocytes and quantified the number of C3-positive astrocytes after stimulation with cytokines. Knockdown of IFT88 tended to suppress primary cilium formation in astrocytes (Fig. S3A–C). Furthermore, IFT88 knockdown tended to decrease the percentage of C3-positive astrocytes, although the cells were stimulated with cytokines (Fig. S3D).

Based on our findings suggesting that the loss of primary cilium in astrocytes suppresses the expression of C3 in reactive astrocytes, we established conditional knockout (cKO) mice. In these mice, LoxP sites were inserted at both ends of exons 4–6 of the IFT88 gene, and cre-dependent deletion of IFT88 was induced by the activation of CrePR recombinase. For this study, we crossed IFT88flox mice with mice expressing CrePR under the control of the astrocyte-specific GLAST promoter (GLAST-CrePR). CrePR activation was induced by RU486, and RU486 was i.p. administered to 8-week-old offspring (Fig. S4A). Following RU486 administration, the phenotype including object recognition and C3 expression in the mice brain was observed (Haycraft et al. 2007; Mishina and Sakimura 2007; Kellendonk et al. 1996) (Fig. 4A).

Fig. 4

Astrocyte-specific IFT88 gene knockout reduces C3 expression. A Timeline of experiments. Eight-week-old mice (IFT88flox/flox; GLAST-CrePR−/− (ctrl) or IFT88flox/flox; GLAST-CrePR± (cKO)) were i.p. injected with 20 nmol/head RU486 twice a week for 2 weeks. Three days after the final injection of RU486, the mice were i.p. injected with 1 mg/kg LPS twice a week. A control group was administered with an equal volume of PBS. Two days after the second LPS injection, an NOR test was performed. B Representative image of Arl13B (green), GFAP (red) and nuclei (Hoechst 33342, blue) in the mouse hippocampal area. Scale bar; 5 μm. The arrow indicates the primary cilium. PBS administration to ctrl mice; n = 5 (biological replicates). LPS administration to ctrl mice; n = 6 (biological replicates). PBS administration to cKO mice; n = 7 (biological replicates). LPS administration to cKO mice; n = 6 (biological replicates). C Percentage of ciliated astrocytes (Arl13B+ GFAP+ cells/GFAP+ cells) in the brain (left, Ctrl + PBS; 15 Arl13B+ GFAP+ cells /119 GFAP+ cells, Ctrl + LPS; 26 Arl13B+ GFAP+ cells /243 GFAP+ cells, cKO + PBS; 14 Arl13B+ GFAP+ cells /338 GFAP+ cells, cKO + LPS; 4 Arl13B+ GFAP+ cells /317 GFAP+ cells), percentage of ciliated GFAP-negative cells (Arl13B+GFAP− cells/GFAP− cells) (middle, Ctrl + PBS; 84 Arl13B+ GFAP− cells /843 GFAP− cells, Ctrl + LPS; 54 Arl13B+ GFAP− cells /873 GFAP− cells, cKO + PBS; 206 Arl13B+ GFAP− cells /1730 GFAP− cells, cKO + LPS; 104 Arl13B+ GFAP− cells /840 GFAP− cells) and cilium length in non-astrocytic cells (right, Ctrl + PBS; 55 cilia, Ctrl + LPS; 45 cilia, cKO + PBS; 51 cilia, cKO + LPS; 57 cilia) shown in Fig. 4B. *P < 0.05, ns, nonsignificant. (Kruskal–Wallis test, Dunn’s multiple comparison). Error bar indicates SD. PBS administration to ctrl mice; n = 5 (biological replicates). LPS administration to ctrl mice; n = 6 (biological replicates). PBS administration to cKO mice; n = 7 (biological replicates). LPS administration to cKO mice; n = 6 (biological replicates). D Representative image of C3 (green), GFAP (red) and nuclei (Hoechst 33342, blue) in the mouse hippocampal area. Scale bar; 20 μm. PBS administration to ctrl mice; n = 5 (biological replicates). LPS administration to ctrl mice; n = 6 (biological replicates). PBS administration to cKO mice; n = 7 (biological replicates). LPS administration to cKO mice; n = 6 (biological replicates). E Percentage of C3-positive astrocytes (C3+GFAP+ cells/GFAP+ cells) shown in Fig. 4D. PBS-treated ctrl mice; n = 5 (biological replicates, 32 C3+ GFAP+/135 GFAP+ cells). LPS-treated ctrl mice; n = 6 (biological replicates, 129 C3+ GFAP+ cells/279 GFAP+ cells). PBS-treated cKO mice; n = 7 (biological replicates, 25 C3+ GFAP+ cells /239 GFAP+ cells). LPS-treated cKO mice; n = 6 (biological replicates, 33 C3+ GFAP+ cells /210 GFAP+ cells). *P < 0.05, ns, nonsignificant. (Kruskal–Wallis test, Dunn’s multiple comparison). Error bar indicates SD. F A representative image of early apoptotic cells detected by the TUNEL method in mouse hippocampal CA1/CA3 regions (left). Scale bar; 20 μm. PBS-treated ctrl mice; n = 6 (biological replicates). LPS-treated ctrl mice; n = 6 (biological replicates). PBS-treated cKO mice; n = 6 (biological replicates). LPS-treated cKO mice; n = 6 (biological replicates). Positive control (PC) slides were used as staining controls. The number of TUNEL-positive cells is shown in the graph (right). *P < 0.05, **P < 0.01, ns, nonsignificant. (Kruskal–Wallis test, Dunn’s multiple comparison). Error bar indicates SD

To initially confirm the efficiency of IFT88 reduction, we quantified the percentage of primary ciliated astrocytes in mouse brain. We first tried to detect IFT88-positive primary cilium in the brain; however, the signal was extremely low even in wild-type mice, so we calculated the percentages by detecting Arl13B, which is commonly used as a primary cilium marker. In RU486-administered IFT88flox/floxGLAST-CrePR (cKO) mouse brains, a tendency of reduction of astrocytic primary cilium was observed (Fig. 4B, C). In contrast, there was no significant difference in the percentage of primary ciliated cells and cilium length in GFAP-negative cells between IFT88flox/+; GLAST-CrePR (ctrl) and cKO mice (Fig. 4C).

These results suggest the astrocyte-specific reduction in the primary cilium in cKO mice. Subsequently, we examined C3 expression in these mice after LPS administration. In control mice, an increase tendency in the number of C3-positive astrocytes was observed upon LPS administration, while no significant increase in C3-positive astrocyte numbers was observed in cKO mice even after LPS administration (Fig. 4D, E). We also examined primary cilium loss in cultured astrocytes. The percentage of primary ciliated astrocytes were comparable between control mouse-derived and C57BL/6 J (wild-type B6J) mouse-derived cultured astrocytes, indicating that it was a feature of only cKO mouse-derived astrocytes treated with RU486 (Fig. S4B, C, E, F). Furthermore, the increase tendency in C3-positive astrocyte numbers induced by LPS stimulation was not observed in cKO mouse-derived astrocytes (Fig. S4G, H).

It has been reported that C3-positive reactive astrocytes induce oligodendrocyte death by secreting saturated lipids (Liddelow et al. 2017; Guttenplan et al. 2021). To examine whether C3-positive reactive astrocytes induce cell death in the brain, we detected apoptotic cells upon LPS administration using the TdT-mediated biotin-dUTP nick end labelling (TUNEL) method. In control mice, LPS administration led to an increase in the proportion of TUNEL-positive cells near the hippocampal CA1/CA3 regions, whereas no significant increase in the proportion of TUNEL-positive cells was observed in cKO mice (Fig. 4F). These results suggest that the functional inhibition of the primary cilium in astrocytes reduced C3 expression in reactive astrocyte.

The Specific Inhibition of Primary Cilium Formation in Astrocytes Ameliorates LPS-Induced Cognitive Impairment

It has been reported that inflammation model mice exhibit a decline in cognitive function (Alzahrani et al. 2022; Borikar et al. 2022; Skrzypczak-Wiercioch and Sałat 2022). Therefore, we investigated mouse cognitive function using the novel object recognition (NOR) test for evaluating long-term memory, the open field (OF) test for determining locomotor activity, and the Y-maze test for assessing short-term memory (Fig. 5A). In the inflammation model mice, the preference for the novel object was significantly lower than that of the control mice, whereas no significant differences were observed in the OF test or Y-maze test (Fig. 5B). Furthermore, cognitive tests were conducted on control and cKO mice. In control mice, the preference for the novel object was reduced upon LPS administration, whereas a decline in the preference for the novel object was tended to attenuate even after LPS administration in cKO mice (Fig. 5C). Compared to wild-type, cKO mice without RU486 administration did not show any differences in cognitive function (Fig. S4D). These results suggest that astrocyte-specific loss of primary cilium attenuated cognitive decline by reducing C3-positive reactive astrocytes in the brain.

Fig. 5

The specific inhibition of primary cilium formation in astrocytes ameliorates LPS-induced cognitive impairment. A Illustration of the mouse behavioural test. The novel object recognition (NOR) test (left), open field (OF) test (middle) and Y-maze test (right) were performed. Figures were created with BioRender.com. B Results of behavioural (NOR; left, OF; middle, Y-maze; right) studies in 3-month-old PBS- and LPS-injected C57BL/6 J mice. PBS-treated mice; n = 6 (biological replicates). LPS-treated mice; n = 6 (biological replicates). **P < 0.01, ns, nonsignificant. (Mann‒Whitney U test). Error bar indicates SD. C Results of ORT studies in cKO mice. PBS-treated ctrl mice; n = 9 (biological replicates). LPS-treated ctrl mice; n=6 (biological replicates). PBS-treated cKO mice; n=6 (biological replicates). LPS-treated cKO mice; n=7 (biological replicates). *P < <0.05, **P < 0.01, ns, nonsignificant. (Kruskal–Wallis test, Dunn’s multiple comparison). Error bar indicates SD